Research map / Learning theory and algorithms / Kernels and regression theory
Deep Nonparametric Quantile Regression under Covariate Shift
Xingdong Feng, Xin He, Yuling Jiao, Lican Kang, Caixing Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use importance weights (WDQR) instead of unweighted DQR | deep quantile regression · univariate model (13), unbounded density ratio, tau=0.5, n=2048 | L1 error (x1e-1): 0.402 -> 0.374 | Improved |
| use importance weights (WDQR) instead of unweighted DQR | deep quantile regression · univariate model (13), unbounded density ratio, tau=0.05, n=2048 | squared L2 error (x1e-1): 0.185 -> 0.100 | Improved |
| use importance weights (WDQR) instead of DQR | deep quantile regression · multivariate model (14), unbounded density ratio, tau=0.5, n=2048 | L1 error: 0.209 -> 0.146 | Improved |
| use importance weights (WDQR) instead of DQR | deep quantile regression · multivariate model (14), unbounded density ratio, tau=0.95, n=2048 | squared L2 error: 0.243 -> 0.135 | Improved |
| use importance weights (WDQR) instead of DQR | deep quantile regression · univariate model (13), bounded density ratio, tau=0.5, n=2048 | L1 error (x1e-1): 0.129 -> 0.128 | No clear effect |
| use importance weights (WDQR) instead of DQR | deep quantile regression · multivariate model (14), bounded density ratio, tau=0.5, n=2048 | L1 error: 0.078 -> 0.076 | No clear effect |
| pre-train the weight function (PWDQR) instead of using true weights (WDQR) | weighted deep quantile regression · univariate model (13), unbounded density ratio, tau=0.5, n=2048 | L1 error (x1e-1): 0.374 -> 0.396 | No clear effect |
| use pre-trained weights (PWDQR) instead of DQR | deep quantile regression · multivariate model (14), unbounded density ratio, tau=0.5, n=2048 | squared L2 error: 0.122 -> 0.055 | Improved |
| increase sample size from 512 to 2048 | DQR · univariate model (13), unbounded density ratio, tau=0.5 | L1 error (x1e-1): 0.912 -> 0.402 | Improved |
Closest papers by meaning
- Quantile regression with ReLU Networks: Estimators and minimax rates (JMLR 2022 · similarity 0.91)
- Sharp Generalization for Nonparametric Regression by Over-Parameterized Neural Networks: A Distribution-Free Analysis in Spherical Covariate (ICML 2025 · similarity 0.87)
- Towards Self-Supervised Covariance Estimation in Deep Heteroscedastic Regression (ICLR 2025 · similarity 0.85)
- Optimizing importance weighting in the presence of sub-population shifts (ICLR 2025 · similarity 0.85)
- Simulation-Based Inference with Quantile Regression (ICML 2024 · similarity 0.85)
- ReTaSA: A Nonparametric Functional Estimation Approach for Addressing Continuous Target Shift (ICLR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
